EIP: Rates, Ratios, Risks - Week 11 Flashcards
Do cross-sectional studies imply cause and effect?
No
Can cross-sectional studies tell us what factors are causing a health problem?
No
What can cross-sectional studies describe?
‘state of play’
In comparison to cross-sectional studies, how would you describe a cohort study?
longitudinal, unlike cross-sectional
What type of study shows us a ‘snapshot’: cross-sectional or cohort?
cross-sectional
What is an odds ratio?
A comparison of the odds of something in 2 different groups
- it gives us the odds of something being true in one group compared to another
Are odds the same as probabilities?
No. Therefore not the same as relative risk
How frequently are odds and risks similar?
often
How would you find the relative risk of something?
by finding the proportion in a population (it’s a probability so divide by total and there you go)
Let’s assume you find 10 myopes in a sample of 45. What are the ODDS of any one individual being a myope?
10:35 (or 10 myopes to 35 non-myopes, or 0.29 to one)
Let’s assume you find 10 myopes in a sample of 45. What is the RISK of any one individual being a myope?
10/45 = 0.22 (a probability)
Which is more important to use: Odds Ratio or Relative Risk? (2)
Depends on the study design
- odds ratio should be used in cross-sectional studies
Are case control studies retrospective or prospective?
retrospective
- you start with the state under investigation already present and go back to try and identify the cause
Can case control studies answer questions about causation?
No. They are too prone to bias
Are case control studies useful for hypothesis generation?
yes
Define Control Event Rate (CER)
The probability of developing an outcome for the control group
CER = c/(c+d)
Define Experimental Event Rate (EER)
The probability of developing an outcome for the experimental (treatment) group
EER = a/(a+b)
Define Relative Risk (RR)
The ratio of the probability of developing an outcome among those receiving the treatment of interest or exposed to a risk factor, compared with the probability of developing the outcome if the intervention or risk factor is present
RR = EER/CER
Define Relative Risk Reduction (RRR)
The extent to which a treatment reduces a risk, in comparison with patients not receiving the treatment of interest
RRR = (CER - EER)/CER
Define Absolute Risk Reduction (ARR)
The difference in the rates of adverse events between experimental and control groups
ARR = CER - EER
Define Number needed to treat (NNT)
The number of patients who must be exposed to an intervention before the outcome of interest occurred (e.g. the number of patients needed to treat to prevent one adverse outcomes)
NNT = 1/ARR
